Information Processing Apparatus, Method and Program

1. An information processing apparatus for managing long-term preference attributes of a user, comprising: a processor; a memory storing the long-term preference attributes, the long-term preference attributes being based on attributes of content selected by the user during a first period; means for receiving the user's selection of content during a second period that is shorter than the first period; means for generating short-term preference attributes of the user from attributes of the content selected by the user during the second period; and decision means for determining content to be recommended from content available to the user, comprising: means for computing a first similarity value based on a similarity between the long-term preference attributes and the short-term preference attributes, the first similarity value indicating how similar the long-term preference attributes are to the short-term preference attributes; means for determining whether the first similarity value is below a reference value; and means for, when it is determined that the first similarity value is below the reference value: generating weight information for recommending the content to the user based on performing a mathematical operation on the first similarity value and a predetermined weight value weighting attributes of the short-term preference attributes that are dissimilar to corresponding attributes of the long-term preference attributes more heavily than attributes of the short-term preference attributes that are similar to corresponding attributes of the long-term preference attributes; computing a second similarity value based on a similarity between the short-term preference attributes of the user and corresponding attributes of the content available to the user and on the weight information; and selecting for recommendation the content from the content available to the user based on the second similarity value such that the selected content tends to have attributes that are dissimilar to corresponding attributes of the long-term preference attributes; and means for recommending the selected content to the user, wherein as the user selects content during the second period, the reference value is reduced.

2. The information processing apparatus according to claim 1 , wherein, if content other than the recommended content is selected by the user, the decision means excludes, from the recommended content, content having attributes dissimilar to attributes of the selected content and one of the long-term preference attributes and the short-term preference attributes.

3. The information processing apparatus according to claim 1 , further comprising: means for communicating with a server storing the content available to the user; and means for transmitting the short-term preference attributes and the weight information to the server via the means for communicating if a difference between the long-term preference attributes and the short-term preference attributes is large and for acquiring, from the server, the computed second similarity value, wherein the means for determining content determines the content to be recommended to the user based on the acquired second similarity value.

4. An information processing method for managing long-term preference attributes of a user, the method comprising: generating the long-term preference attributes of the user from attributes of content selected by the user during a first period; receiving the user's selection of content during a second period that is shorter than the first period; generating short-term preference attributes of the user from attributes of the content selected by the user during the second period; and determining content to be recommended from content available to the user by: computing a first similarity value based on a similarity between the long-term preference attributes and the short-term preference attributes, the first similarity value indicating how similar the long-term preference attributes are to the short-term preference attributes; determining whether the first similarity value is below a reference value; and when it is determined that the first similarity value is below the reference value: generating weight information for recommending the content to the user based on performing a mathematical operation on the first similarity value and a predetermined weight value weighting attributes of the short-term preference attributes that are dissimilar to corresponding attributes of the long-term preference attributes more heavily than attributes of the short-term preference attributes that are similar to corresponding attributes of the long-term preference attributes; computing a second similarity value based on a similarity between the short-term preference attributes of the user and corresponding attributes of the content available to the user and on the weight information; and selecting for recommendation the content from the content available to the user based on the second similarity value such that the selected content tends to have attributes that are dissimilar to corresponding attributes of the long-term preference attributes; and recommending the selected content to the user, wherein as the user selects content during the second period, the reference value is reduced.

5. A computer-readable memory device storing a program which, when executed by a processor, causes the processor to perform a method for controlling an information processing apparatus for managing long-term preference attributes of a user, the method comprising: generating the long-term preference attributes of the user from attributes of content selected by the user during a first period; receiving the user's selection of content during a second period that is shorter than the first period; generating short-term preference attributes of the user from attributes of the content selected by the user during the second period; and determining content to be recommended from content available to the user by: computing a first similarity value based on a similarity between the long-term preference attributes and the short-term preference attributes, the first similarity value indicating how similar the long-term preference attributes are to the short-term preference attributes; determining whether the first similarity value is below a reference value; and when it is determined that the first similarity value is below the reference value: generating weight information for recommending the content to the user based on performing a mathematical operation on the first similarity value and a predetermined weight value weighting attributes of the short-term preference attributes that are dissimilar to corresponding attributes of the long-term preference attributes more heavily than attributes of the short-term preference attributes that are similar to corresponding attributes of the long-term preference attributes; computing a second similarity value based on a similarity between the short-term preference attributes of the user and corresponding attributes of the content available to the user and on the weight information; and selecting for recommendation the content from the content available to the user based on the second similarity value such that the selected content tends to have attributes that are dissimilar to corresponding attributes of the long-term preference attributes; and recommending the selected content to the user, wherein as the user selects content during the second period, the reference value is reduced.

6. An information processing apparatus for managing long-term preference attributes of a user and for controlling a device, the information processing apparatus comprising: a processor; a memory storing the long-term preference attributes of the user, the long-term preference attributes being based on attributes of content selected by the user during a first period; means for receiving the user's selection of content during a second period that is shorter than the first period; means for generating short-term preference attributes of the user from attributes of the content selected by the user during the second period; and means for controlling the device to recommend content from content available to the user, comprising: means for computing a first similarity value based on a similarity between the long-term preference attributes of the user and the short-term preference attributes of the user, the first similarity value indicating how similar the long-term preference attributes are to the short-term preference attributes; means for determining whether the first similarity value is below a reference value; and means for, when it is determined that the first similarity value is below the reference value: generating weight information for recommending the content to the user based on performing a mathematical operation on the first similarity value and a predetermined weight value weighting attributes of the short-term preference attributes that are dissimilar to corresponding attributes of the long-term preference attributes more heavily than attributes of the short-term preference attributes that are similar to corresponding attributes of the long-term preference attributes; computing a second similarity value based on a similarity between the short-term preference attributes of the user and corresponding attributes of the content available to the user and on the weight information; and selecting for recommendation the content from the content available to the user based on the second similarity value such that the selected content tends to have attributes that are dissimilar to corresponding attributes of the long-term preference attributes; and means for recommending the selected content to the user, wherein as the user selects content during the second period, the reference value is reduced.

7. The information processing apparatus according to claim 6 , wherein the means for controlling the device includes means for estimating a mood of the user based on the weight information.

8. The information processing apparatus according to claim 7 , wherein the control means notifies the means for recommending of the estimated mood; and the means for recommending recommends content corresponding to the estimated mood.

9. An information processing method for an information processing apparatus for managing long-term preference attributes of a user and for controlling a device, the method comprising: generating the long-term preference attributes of the user from attributes of content selected by the user during a first period; receiving the user's selection of content during a second period that is shorter than the first period; generating short-term preference attributes of the user from attributes of the content selected by the user during the second period; and controlling the device to recommend content from content available to the user, the controlling including: computing a first similarity value based on a similarity between the long-term preference attributes and the short-term preference attributes, the first similarity value indicating how similar the long-term preference attributes are to the short-term preference attributes; determining whether the first similarity value is below a reference value; and when it is determined that the first similarity value is below the reference value: generating weight information for recommending the content to the user based on performing a mathematical operation on the first similarity value and a predetermined weight value weighting attributes of the short-term preference attributes that are dissimilar to corresponding attributes of the long-term preference attributes more heavily than attributes of the short-term preference attributes that are similar to corresponding attributes of the long-term preference attributes; computing a second similarity value based on a similarity between the short-term preference attributes of the user and corresponding attributes of the content available to the user and on the weight information; and selecting for recommendation the content from the content available to the user based on the second similarity value such that the selected content tends to have attributes that are dissimilar to corresponding attributes of the long-term preference attributes of the user; recommending the selected content to the user, and reducing the reference value as the user selects content during the second period.

10. A computer-readable memory device storing a program which, when executed by a processor, causes the processor to perform a method for managing long-term preference attributes of a user and for controller a device, the method comprising: generating the long-term preference attributes of the user from attributes of content selected by the user during a first period; receiving the user's selection of content during a second period that is shorter than the first period; generating short-term preference attributes of the user from attributes of the content selected by the user during the second period; and controlling the device to recommend content from content available to the user, the controlling comprising: computing a first similarity value based on a similarity between the long-term preference attributes and the short-term preference attributes, the first similarity value indicating how similar the long-term preference attributes are to the short-term preference attributes; determining whether the first similarity value is below a reference value; and when it is determined that the first similarity value is below the reference value: generating weight information for recommending the content to the user based on performing a mathematical operation on the first similarity value and a predetermined weight value weighting attributes of the short-term preference attributes that are dissimilar to corresponding attributes of the long-term preference attributes more heavily than attributes of the short-term preference attributes that are similar to corresponding attributes of the long-term preference attributes; computing a second similarity value based on a similarity between the short-term preference attributes of the user and corresponding attributes of the content available to the user and on the weight information; and selecting for recommendation the content from the content available to the user based on the second similarity value such that the selected content tends to have attributes that are dissimilar to corresponding attributes of the long-term preference attributes of the user; recommending the selected content to the user, and reducing the reference value as the user selects content during the second period.

11. An information processing apparatus for managing long-term preference attributes of a user, comprising: a memory storing: a long-term preference attributes generation section for generating the long-term preference attributes of the user from attributes of content selected by the user during a first period; a user input section for receiving the user's selection of content during a second period that is shorter than the first period; a short-term preference attributes generation section for generating short-term preference attributes of the user from attributes of content selected by the user during the second period; and a control section for controlling the device to recommend content from content available to the user, comprising: a similarity computing section for computing a first similarity value based on a similarity between the long-term preference attributes and the short-term preference attributes, the first similarity value indicating how similar the long-term preference attributes are to the short-term preference attributes; a determination section for determining whether the first similarity value is below a reference value; and a recommendation control section for, when it is determined that the first similarity value is below the reference value: generating weight information for recommending the content to the user based on performing a mathematical operation on the first similarity value and a predetermined weight value weighting attributes of the short-term preference attributes that are dissimilar to corresponding attributes of the long-term preference attributes more heavily than attributes of the short-term preference attributes that are similar to corresponding attributes of the long-term preference attributes; computing a second similarity value based on a similarity between the short-term preference attributes of the user and corresponding attributes of the content available to the user and on the weight information; and selecting for recommendation the content from the content available to the user based on the second similarity value such that the selected content tends to have attributes that are dissimilar to corresponding attributes of the long-term preference attributes; and recommending the selected content to the user; and a processor configured to execute the sections stored in the memory, wherein as the user selects content during the second period, the reference value is reduced.

12. An information processing apparatus for managing long-term preference attributes of a user and for controlling a device, comprising: a memory storing: a long-term preference attributes generation section for generating the long-term preference attributes of the user from attributes of content selected by the user during a first period; a user input section for receiving the user's selection of content during a second period that is shorter than the first period; a short-term preference attributes generation section for generating short-term preference attributes of the user from attributes of content selected by the user during the second period; and a control section for controlling the device to recommend content from content available to the user, comprising: a similarity computing section for computing a first similarity value based on a similarity between the long-term preference attributes of the user and the short-term preference attributes of the user, the first similarity value indicating how similar the long-term preference attributes are to the short-term preference attributes; a determination section for determining whether the first similarity value is below a reference value; and a recommendation control section for, when it is determined that the first similarity value is below the reference value: generating weight information for recommending content to the user based on performing a mathematical operation on the first similarity value and a predetermined weight value weighting attributes of the short-term preference attributes that are dissimilar to corresponding attributes of the long-term preference attributes more heavily than attributes of the short-term preference attributes that are similar to corresponding attributes of the long-term preference attributes; computing a second similarity value based on a similarity between the short-term preference attributes and corresponding attributes of the content available to the user and on the weight information; selecting for recommendation the content from the content available to the user based on the second similarity value such that the selected content tends to have attributes that are dissimilar to corresponding attributes of the long-term preference attributes; and recommending the selected content to the user; and a processor configured to execute the sections stored in the memory, wherein as the user selects content during the second period, the reference value is reduced.